Fuzzy Bio- Bio-interface: Can fuzzy sets sets be an interface with brain? Isao Hayashi † and Suguru N. Kudoh ‡ † Faculty of Informatics, Kansai University, Takatsuki, Osaka 569-1095, JAPAN ‡ School of Science and Technology, Kwansei Gakuin University, Sanda, Hyogo 669-1337, JAPAN two kinds of functions: (1) a decoding of the response action potentials to the control signal of outside machine and computer, and (2) an encoding of the sensor signal of the outside machine and computer to pattern of stimuli in brain and neuronal networks. Unfortunately, it is very difficult to identify such a function for the interface between machine and living brain and neuronal networks. Here we consider such an interface within the framework of fuzzy system. As a result, our study is supportive of this framework as a strong tool of the bio-interface. During the Japanese fuzzy boom in 1990's, fuzzy logic has been proven effective to translate human experience and sensitivity into control signals of machines. Tsukamoto[3] has argued a concept of fuzzy interface such that fuzzy sets is regarded as a useful tool to intermediate between language and mathematics. We believe that the ABSTRACT framework of fuzzy system is essential for BCI Recently, many attractive brain-computer and BMI, thus name this technology “fuzzy interface and brain-machine interface have been bio-interface.” proposed[1,2]. The outer computer and machine In this lecture, we introduce a fuzzy are controlled by brain action potentials detected bio-interface between a culture dish of rat through a device such as near-infrared hippocampal neurons and the khepera robot. We spectroscopy (NIRS) and electroencephalograph propose a model to analyze logic of signals and (EEG), and some discriminant model determines connectivity of electrodes in a culture dish[4], and a control process. However, under the condition show the bio-robot hybrid we developed[5,6]. Rat where spontaneous action-potentials and hippocampal neurons are organized into complex evoked-action potentials are contained in brain networks in a culture dish with 64 planar signal asynchronously, we need a model that microelectrodes. A multi-site recording system for serves as an interface between brain and machine extracellular action potentials is used in order to for a better stable control in order to prevent record their activities in living neuronal networks runaway reaction of machine. This interface plays and to supply input from the outer world to the a very important role to secure the stability of vitro living neural networks. The living neuronal outer computer and machine. The interface has networks are able to express several patterns independently, and such patterns represent connectives, that consist of both t-norms and fundamental mechanisms for intelligent t-conorms[9,10], in order to analyze those three information processing[7]. electrodes (Figure 1). We have obtained the First, we discuss how to indicate the experimental result such that the parameter(s) of logicality and connectivity from living neuronal fuzzy connectives become infinity. Given this network in vitro. We follow the works of result, we conclude that a pulse at the 60th Bettencourt et al.[8] such that they classify the channel (60el) propagates to the spreading area: connectivity of action potentials of three (51el, 59el), (43el, 50el) and (35el, 42el); and that electrodes on multi-site recording system the logic of signals among the electrodes was according to their entropies and have discussed shifted to the logical sum from the drastic product. the characteristic of each classification. However, Consequently, the logic of signals among they only discuss the static aspects of connectivity electrodes drastically changes from the strong relations among the electrodes but not the AND-relation to the weak OR-relation when a dynamics of such connectivity concerning how the crowd of the pulses was fired. strength of electrode connection changes when a spike is fired. To address this issue, we develop a new algorithm using parametric fuzzy Figure 1: Algorithm to Analyze Action Potentials in Cultured Neuronal Network Next, to control a robot, several characteristics of the living neuronal networks are represented as fuzzy IF-THEN rules. There are many works of robots that are controlled by the responses from living neuronal networks[11-15]. Unfortunately, they have not yet achieved a certain task that experimenter desired. We show a robot system that controlled by a living neuronal network through the fuzzy bio-interface in order to achieve such a task (Figure 2). This fuzzy bio-interface consists of two sets of fuzzy IF-THEN rules: (1) to translate sensor signals of robot into stimuli for the living neuronal network, and (2) to control (i.e. to determine the action of) robot based on the responses from the living neuronal network. We estimated the learning of living neuronal networks with an example of straight running with neuro-robot hybrid. Among 20 trials, the robot completed the task 16 times, and it crashed on the wall and stopped there 4 times. In this result, we may conclude that the logic of signals among living neuronal networks represented as Figure 2: Living Neuronal Network and Robot fuzzy IF-THEN rules for the fuzzy bio-interface is rather efficient and effective comparing to the other similar works. In such works, the success rate of 80% is considered extremely high. Ensemble Adaptation to Represent Velocity of an Artificial Actuator Controlled by a ACKNOLEDGEMENT Brain-machine Interface, Journal of I would like to express my gratitude to my Neuroscience, Vol.25, No.19, pp.4681-4693, collaborators, Megumi Kiyotoki, Kansai 2005. University, Japan and Ai Kiyohara, Minori [2] L.R.Hochberg, M.D.Serruya, G.M.Friehs, Tokuda, Kwansei Gakuin University, Japan. This J.A.Mukand, M.Saleh, A.H.Caplan, work is partially supported by the Ministry of A.Branner, D.Chen, R.D.Penn, Education, Culture, Sports, Science, and J.P.Donoghue: Neuronal Ensemble Control of Technology of Japan under Grant-in-Aid for Prosthetc Devices by a Human with Scientific Research 18500181, 19200018, and Tetraplegia, Nature, Vol.442, pp.164-173, 18048043 and by the Organization for Research 2006. and Development of Innovative Science and [3] Y.Tsukamoto: Fuzzy Sets as an Interface Technology (ORDIST) of Kansai University. between Language Model and Mathmatics Model, Proc. of the 24th Fuzzy System REFERENCES Symposimu, Vol.251-254, 2008. [1] M.A.Lebedev, J.M.Carmera, J.E.O'Doherty, [4] I.Hayashi, M.Kiyotoki, A.Kiyohara, M.Zacksenhouse, C.S.Henriquez, M.Tokuda, and S.N.Kudoh: Acquisition of J.C.Principe, and M.A.L.Nicolelis: Cortical Logicality in Living Neuronal Networks and its Operation to Fuzzy Bio-Robot System, Flight Control with Living Neuronal Proc. of 2010 IEEE International Conference Networks on Microelectrode Arrays, Proc. of on Fuzzy Systems (FUZZ-IEEE2010) in 2010 2005 IEEE International Joint Conference on IEEE World Congress on Computational Neural Networks (IJCNN2005), Intelligence (WCCI2010), pp.543-549, 2010. pp.1549-1551, Montreal, Canada , 2005. [5] S.N.Kudoh, I.Hayashi, and T.Taguchi: [13] Z.C.Chao, D.J.Bakkum, and S.M. Potter: Synaptic Potentiation Re-organized Shaping Embodied Neural Networks for Functional Connections in a Cultured Adaptive Goal-directed Behavior, PLoS Neuronal Network Connected to a Moving Comput Biol, Vol.4, No.3, e1000042, 2008. Robot, Proc. of the 5th International Meeting [14] P.Marks: Rat-brained Robots take Their on Substrate-Integrated Micro Electrode First Steps, New Scientist, Vol.199, No.2669, Arrays (MEA2006), pp.51-52, Reutlingen, pp.22-23, 2008. Germany, 2006. [15] K.Warwick: Implications and Consequences [6] S.N.Kudoh, C.Hosokawa, A.Kiyohara, of Robots with Biological Brains, Journal of T.Taguchi, and I.Hayashi: Biomodeling Ethics and Information Technology, Vol.12, System - Interaction between Living No.3, pp.223-234, 2010. Neuronal Network and Outer World, Journal of Robotics and Mechatronics, Vol.19, No.5, BIOGRAPHICAL SKETCH pp.592-600, 2007. Isao Hayashi is Professor of Informatics at [7] S.N.Kudoh and T.Taguchi: Operation of Kansai University, Japan. After he received his Spatiotemporal Patterns Stored in Living B.Eng. degree in Industrial Engineering from Neuronal Networks Cultured on a Osaka Prefecture University, he worked at Sharp Microelectrode Advanced Array, Corporation, Japan. After he received his M.Eng. Computational Intelligence and Intelligent degree from Osaka Prefecture University in 1987, Informatics, Vol.8, No2, pp.100-107, 2003. he was a Senior Research Fellow of the Central [8] L.M.A.Bettencourt, G.J.Stephens, M.I.Ham, Research Laboratory of Matsushita Electric and G.W.Gross: Functional Structure of Industrial (Panasonic) Co. Ltd and proposed a Cortical Neuronal Networks Grown in Vitro, neuro-fuzzy system on intelligent control. Phisical Review, Vol.75, p.02915, 2007. He received his D.Eng. degree based on [9] B.Schweizer and A.Sklar: Associative his contributions to the neuro-fuzzy model from Functions and Statistical Triangle Osaka Prefecture University in 1991. He then Inequalities, Publicationes Mathematicae joined Faculty of Management Information of Debrecen, Vol.8, pp.169-186, 1961. Hannan University in 1993 and joined Faculty of [10] I.Hayashi, E.Naito, and N.Wakami: Proposal Informatics of Kansai University in 2004. He is for Fuzzy Connectives with a Learning an editorial member of International Journal of Function Using the Steepest Descent Hybrid Intelligent Systems, Journal of Advanced Method, Japanese Journal of Fuzzy Theory Computational Intelligence and Intelligent and Systems, Vol.5, No.5, pp.705-717, 1993. Informatics, and has served on many conference [11] D.J.Bakkum, A.C.Shkolnik, G.Ben-Ary, program and organizing committees. He is the P.Gamblen, T.B.DeMarse, and S.M. Potter: president of Kansai Chapter of Japan Society for Removing Some `A' from AI: Embodied Fuzzy Theory and Intelligent Informatics (SOFT), Cultured Networks, in Embodied Artificial and the chair of the Technical Group on Brain Intelligence, editered by F.Iida, R.Pfeifer, and Perception in SOFT. He research interests L.Steels, and Y.Kuniyoshi, New York, include visual models, neural networks, fuzzy Springer, pp.130-145, 2004. systems, neuro-fuzzy systems, and [12] T.B.DeMarse and K.P.Dockendorf: Adaptive brain-computer interface. Suguru N. Kudoh received his Master‘s degree in Biophysical Engineering in 1995 and PhD from the Osaka university in 1998. He was a research fellow of JST(Japan science and technology agency) from 1997 to 1998, and a research scientist of National Institute of Advanced Industrial Science and Technology (AIST) from 1998 to 2009. Now he is an associate professor at Kwansei Gakuin university. The aim of his research is to elucidate relationship between dynamics of neuronal network and brain information processing. He analyses spatio-temporal pattern of electrical activity in rat hippocampal cells cultured on multi-electrode arrays or acute slice of basal ganglia. He is also developing Bio-robotics hybrid system in which a living neuronal network is connected to a robot body via control rules, corresponding to agenetically provided interfaces between a brain and a peripheral system. He believes that mind emerges from fluctuation of dynamics in hierarchized interactions between cells.